JHU Ref. No. C18037_P18037-02 Atty. Docket No. JHU-42990.601 MULTIPLEXED DIFFERENTIAL ANALYSIS OF PROTEIN-PROTEIN INTERACTIONS IN CANCER CELLS AND SINGLE CELLS CROSS-REFERENCE TO RELATED APPLICATION This application claims the benefit of U.S. Provisional Patent Application No.63/568,749, filed March 22, 2024, which is incorporated by reference herein in its entirety. FIELD The present disclosure provides methods and systems for identifying protein interactions from bulk and single cell samples. BACKGROUND Cancer arises from changes to gene and protein activities and dysregulation of biological pathways. Pathways are built by mechanistic physical interactions. Methods to measure these interactions are limited. Yeast two-hybrid and related methods involve highly engineered systems that are removed from the cellular context of actual tumor cells. Other cell-based technologies require expression of proteins modified to include tags for affinity purification, proximity labeling, or imaging, which are feasible for applications to model systems but not to tumor specimens obtained from human subjects. Methods that can be multiplexed, that can identify quantitative differences in interaction partners, and that can dissect heterogeneity of human tumors at the level of single tumor cells are lacking. Measuring the state of the transcriptome in bulk through RNA-seq, and dissecting heterogeneity through single-cell RNA-seq have transformed the ability to understand transcriptional changes in cancer, but still do not directly reveal the aberrant interactions that cause these changes. SUMMARY Disclosed herein are methods for detecting protein interactions in a sample. In some embodiments, the methods comprise: (i) contacting the sample with at least one first detection agent, wherein each of the at least one first detection agent comprises an antibody, or fragment thereof, to a first protein of interest covalently attached to a first barcode oligonucleotide; (ii) contacting the sample with at least one second detection agent, wherein the second detection agent comprises an antibody, or fragment thereof, to a second protein of interest covalently attached to a second barcode oligonucleotide, wherein the first barcode oligonucleotide and the second barcode oligonucleotide are configured for ligation to each other when in close proximity; (iii) ligating the first barcode oligonucleotide and the second barcode oligonucleotide to form a contiguous oligonucleotide; and (iv) detecting formation of the contiguous oligonucleotide.
JHU Ref. No. C18037_P18037-02 Atty. Docket No. JHU-42990.601 In some embodiments, the methods comprise: (i) contacting the sample with at least one first detection agent, wherein each of the at least one first detection agent comprises an antibody, or fragment thereof, to a first protein of interest covalently attached to a first barcode oligonucleotide comprising a first cleavage site; (ii) contacting the sample with at least one second detection agent, wherein the second detection agent comprises an antibody, or fragment thereof, to a second protein of interest covalently attached to a second barcode oligonucleotide comprising a second cleavage site, wherein the second cleavage site and the first cleavage site create complementary single strand ends; (iii) contacting the sample with: one or more cleavage agents configured to act on the first cleavage site and second cleavage site, and a ligase; and (iv) detecting formation of a contiguous oligonucleotide comprising the first barcode oligonucleotide and the second barcode oligonucleotide. In some embodiments, detecting formation of the contiguous oligonucleotide indicates that the first protein of interest and the second target protein of interest are in close proximity. In some embodiments, the method comprises contacting the sample with a plurality of first detection agents, each directed to a different first protein of interest, and/or contacting the sample with a plurality of second detection agents, each directed to a different second protein of interest. In some embodiments, step (i) and step (ii) are performed simultaneously; step (i) is performed before step (ii); or step (ii) is performed before step (i). In some embodiments, step (i) and/or step (ii) comprises contacting the sample with the first detection agent and/or second detection agent for about 10 minutes to about 48 hours. In some embodiments, one of the first detection agent or second detection agent is bound to a solid surface or comprises a functional group configured to bind to a solid surface. In some embodiments, the solid surface is a bead or particle. In some embodiments, the method further comprises separating or isolating solid surface bound detection agents and binding partners from the sample. In some embodiments, the first barcode oligonucleotide and the second barcode oligonucleotide have a length of about 20 to 150 basepairs. In some embodiments, the first barcode oligonucleotide and the second barcode oligonucleotide each comprise a barcode and a unique molecular identifier (UMI) flanked by primer binding sites. In some embodiments, each of the first and second cleavage sites are distal to the barcode and UMI in relationship to the site of attachment to the antibody, or fragment thereof. In some embodiments, detecting formation of a contiguous oligonucleotide comprises: isolating contiguous oligonucleotides from the sample; and amplifying and/or sequencing contiguous oligonucleotides.
JHU Ref. No. C18037_P18037-02 Atty. Docket No. JHU-42990.601 In some embodiments, the methods further comprise mapping the contiguous oligonucleotides to the protein interactions based on the barcode and/or UMI. In some embodiments, the sample is a biological sample. Also disclosed herein are systems for detecting protein interactions in a sample. In some embodiments, the systems comprise: at least one first detection agent, wherein each of the at least one first detection agent comprises an antibody, or fragment thereof, to a first protein of interest covalently attached to a first barcode oligonucleotide; and at least one second detection agent, wherein the second detection agent comprises an antibody, or fragment thereof, to a second protein of interest covalently attached to a second barcode oligonucleotide. In some embodiments, the first barcode oligonucleotide and the second barcode oligonucleotide are configured for ligation to each other when in close proximity. In some embodiments, the first and second barcode oligonucleotides are double stranded. In some embodiments, the first barcode oligonucleotide and the second barcode oligonucleotide comprise a barcode and a unique molecular identifier (UMI) flanked by primer binding sites. In some embodiments, the first barcode oligonucleotide and the second barcode oligonucleotide each contain a cleavage site configured to create complementary single strand ends. In some embodiments, the cleavage site is distal to the barcode and UMI in relationship to the site of attachment to the antibody, or fragment thereof. In some embodiments, the system comprises a plurality of first detection agents, each directed to a different first protein of interest. In some embodiments, the system comprises a plurality of second detection agents, each directed to a different second protein of interest. In some embodiments, one of the first detection agent or second detection agent is bound to a solid surface or comprises a functional group configured to bind to a solid surface. In some embodiments, the solid surface is a bead or particle. In some embodiments, the system further comprises one or more primers configured to bind to the primer binding sites on the first barcode oligonucleotide and/or the second barcode oligonucleotide. Other aspects and embodiments of the disclosure will be apparent in light of the following detailed description and accompanying figures. BRIEF DESCRIPTION OF THE DRAWINGS FIG.1 is a table of characteristics and advantages for Multiplexed Detection of Protein- Protein Interactions (MuDPPI) as compared to existing methods - yeast two-hybrid (Y2H), immunoprecipitation western blot (IP-WB), affinity purification-mass spectrometry (AP-MS), and
JHU Ref. No. C18037_P18037-02 Atty. Docket No. JHU-42990.601 proximity-dependent biotin identification (BioID) - for analyzing interactions. MuDPPI facilitates examination of protein interaction in biological samples with endogenous proteins in native cells, rather than labeled proteins in engineered cells. Multiplexed barcodes enable differential analysis and single-cell analysis for many-vs-many pairs. FIG.2 is a schematic of MuDPPI technology for label-free, differential, and single-cell analysis of protein-protein interactions. FIG.3 shows the antibody generation and validation pipeline. The mAb production pipeline includes antigen production; hybridoma production; primary validation using protein arrays; secondary validation by IP, IB, ChIP-seq, and IHC; and, finally, distribution as a community resource. FIGS.4 shows barcoded mAbs. FIG.4A is a schematic of chemical reactions to tether anchor oligos to mAbs. FIG.4B shows the quality control of anchor oligo-conjugated mAbs. FIG. 4C shows the quality control of dsDNA barcoded mAbs. SMCC is a linker (e.g., a heterobifunctional amine-to-sulfhydryl crosslinker) succinimidyl 4-(N-maleimidomethyl)cyclohexane-1-carboxylate. FIG.5 shows proof-of-concept assays to detect hetero- and homodimer formation of transcription factors. The -ΔCt values were obtained by subtracting the GAPDH Ct value from each corresponding TF pair’s Ct. For example, the Ct value of GAPDH-Fos (negative control) was subtracted from those of each TF pair obtained on the anti-Fos immobilized beads to obtain the -ΔCt values. The assays were performed in triplicate to calculate p-values. The obtained -ΔCt values of each TF pairs were also obtained from the anti-Max, -Jun, and -Mad1-immobilized beads. FIG.6 shows the profile of PPI changes during mitotic cell cycle. As expected, the CDK6/CycD1 and CDK1/CycB dimers are found most significantly enhanced with respective p- values of <0.0001 and <0.001 (t-tests based on 2-ΔCt values). FIG.7 shows the detection of transient PPIs in the EGFR pathway. ΔCt values are plotted at different time points post EGF treatment. Each MuDPPI assay was performed in triplicate. Two and three asterisk indicate significant p-values of <0.01 and <0.0001, respectively (t-tests of ΔCt values). FIG.8 shows PPI detection at the single-cell level. Ct values of two known heterodimers (Jun-JunB & Jun-Fos) and one homodimer (JunB-JunB) were significantly lower than those of the GAPDH controls in both single-cell and 50-cell assays. The p-values denoted by one and two asterisks are <0.01 and 0.001, respectively (t-tests of Ct values). FIGS.9A and 9B show MuDPPI analysis of cell cycle in HEK 293T samples down to single cells. FIG.9A is samples analyzed by counts of individual proteins cluster according to cell cycle phase (left), including samples generating from varying numbers of cells (right). FIG.9B is
JHU Ref. No. C18037_P18037-02 Atty. Docket No. JHU-42990.601 samples generated from G1 and S phase cells, with varying number of cells per sample, were analyzed to generate counts of individual protein-protein pairs, which were then used for a PCA embedding. Left panel: Samples cluster according to cell cycle phase, with S in the top left and G1 in the bottom right. Right panel: A second axis shows variation with the number of cells per sample, with large cell numbers in the bottom left and single cells in the top right. DETAILED DESCRIPTION Cancer cells subvert normal developmental and signaling pathways, changing the activities of signal transduction pathways and gene regulatory networks and eliminating native interactions and/or creating new molecular interactions. Because of these dysregulated interactions, there is a need to analyze molecular interactions in cancer cells (e.g., tumor specimens), which may differ from interactions in normal cells. Unfortunately, the most widely available methods for measuring protein- protein interactions at scale, including yeast two-hybrid (Y2H), affinity purification-mass spectrometry (AP-MS), and proximity-dependent biotin identification (BioID), generally require engineered cellular systems or expression of tagged proteins. While Y2H campaigns have generated interaction databases for yeast, fruit, and human, the Y2H system expresses human proteins in yeast. The resulting interactions can be difficult to interpret without knowledge that individual proteins are present in a cell and do not consider the cellular and molecular context and presence of accessory factors, e.g., chaperones, scaffolding proteins, and post-translational modifications. AP-MS and BioID use mass spectrometry to identify proteins co-complexed with a bait protein. In most AP-MS systems and all BioID systems, the bait protein is engineered to contain an affinity tag or functional domain. Just as tables of reference transcriptomes in normal tissues are not sufficient for understand aberrant gene expression in cancer, tables of reference interactomes are an important resource but on their own are limited in generating the information to understand, prevent, and treat cancer. Because changes in protein interactions can be difficult to predict, unbiased methods are fundamental to the analysis. Unfortunately, most quantitative assays for interaction activity are low- throughput and require previous knowledge of exactly the interaction to probe. The most-used method may be immunoprecipitation western blot (IP-WB), which is applicable to native cells but is practically limited to tests of individual protein-vs-protein pairs. Similarly, imaging methods such as fluorescence resonance energy transfer (FRET)require the pair to be specified and engineered with imaging tags. Methods such as AP-MS and BioID requiring expression of tagged proteins provide questionable quantitative evidence for differential interaction analysis. Heterogeneity is a defining characteristic of cancer. Somatic mutations introduce differences between germline and cancer genomes. Clonal heterogeneity arises as different cancer
JHU Ref. No. C18037_P18037-02 Atty. Docket No. JHU-42990.601 cells acquire new mutations, creating differences in the ability to proliferate, metastasize, and resist therapies. Epigenetic, epigenomic, and gene expression differences can create cell state differences in cancer cells that are genetically identical, analogous to stable cell fates in untransformed cells. Single-cell methods for DNA and RNA sequencing have become essential components of cancer research. For proteomics, however, single-cell methods have been more challenging. A fundamental reason is the ease of using PCR or rolling circle to amplify signals from nucleic acids, but no similar methods exist to amplify signals from proteins. Mass spectrometry methods (AP-MS and BioID) require more protein input than is typically present in a single cell. Interactions that are transient or that involve low-abundance proteins are difficult to analyze at all, let alone at the single-cell level. Technologies that enable single-cell analysis would have immediate value for cancer research. The present disclosure provides a platform technology for Multiplexed Detection of Protein-Protein Interactions (MuDPPI) that allows highly multiplexed detection of protein-protein interactions inside cells, tissues, and single cells and enables entirely new capabilities for quantifying the cancer interactome. The platform utilizes barcoded mAbs and converts the detection of protein- protein interactions to sequencing readouts (e.g., Next Generation sequencing) thereby enabling bulk tissue/cell and single-cell studies. The MuDPPI platform provides label-free, differential analysis of protein-protein interactions in cancer cells and single cells not been achieved with any other existing technology (Fig.1). Section headings as used in this section and the entire disclosure herein are merely for organizational purposes and are not intended to be limiting. 1. Definitions The terms “comprise(s),” “include(s),” “having,” “has,” “can,” “contain(s),” and variants thereof, as used herein, are intended to be open-ended transitional phrases, terms, or words that do not preclude the possibility of additional acts or structures. However, two or more copies are also contemplated. The singular forms “a,” “and” and “the” include plural references unless the context clearly dictates otherwise. The present disclosure also contemplates other embodiments “comprising,” “consisting of,” and “consisting essentially of,” the embodiments or elements presented herein, whether explicitly set forth or not. For the recitation of numeric ranges herein, each intervening number there between with the same degree of precision is explicitly contemplated. For example, for the range of 6-9, the numbers 7 and 8 are contemplated in addition to 6 and 9, and for the range 6.0-7.0, the number 6.0, 6.1, 6.2, 6.3, 6.4, 6.5, 6.6, 6.7, 6.8, 6.9, and 7.0 are explicitly contemplated.
JHU Ref. No. C18037_P18037-02 Atty. Docket No. JHU-42990.601 The term “and/or” as used in a phrase such as “A and/or B” herein is intended to include both A and B; A or B; A (alone); and B (alone). Likewise, the term “and/or” as used in a phrase such as “A, B, and/or C” is intended to encompass each of the following embodiments: A, B, and C; A, B, or C; A or C; A or B; B or C; A and C; A and B; B and C; A (alone); B (alone); and C (alone). Unless otherwise defined herein, scientific and technical terms used in connection with the present disclosure shall have the meanings that are commonly understood by those of ordinary skill in the art. The meaning and scope of the terms should be clear; in the event, however of any latent ambiguity, definitions provided herein take precedent over any dictionary or extrinsic definition. Further, unless otherwise required by context, singular terms shall include pluralities and plural terms shall include the singular. “Antibody” and “antibodies” as used herein refers to monoclonal antibodies, monospecific antibodies (e.g., which can either be monoclonal, or may also be produced by other means than producing them from a common germ cell), multi-specific antibodies, human antibodies, humanized antibodies (fully or partially humanized), animal antibodies such as, but not limited to, a bird (for example, a duck or a goose), a shark, a whale, and a mammal, including a non-primate (for example, a cow, a pig, a camel, a llama, a horse, a goat, a rabbit, a sheep, a hamster, a guinea pig, a cat, a dog, a rat, a mouse, etc.) or a non-human primate (for example, a monkey, a chimpanzee, etc.), recombinant antibodies, chimeric antibodies, single-chain Fvs (“scFv”), single chain antibodies, single domain antibodies, Fab fragments, F(ab’) fragments, F(ab’)2 fragments, disulfide-linked Fvs (“sdFv”), and anti-idiotypic (“anti-Id”) antibodies, dual-domain antibodies, dual variable domain (DVD) or triple variable domain (TVD) antibodies (dual-variable domain immunoglobulins and methods for making them are described in Wu, C., et al., Nature Biotechnology, 25(11):1290-1297 (2007) and PCT International Application WO 2001/058956, the contents of each of which are herein incorporated by reference), or domain antibodies (dAbs) (e.g., such as described in Holt et al., Trends in Biotechnology 21:484-490 (2014)), and including single domain antibodies sdAbs that are naturally occurring, e.g., as in cartilaginous fishes and camelid, or which are synthetic, e.g., nanobodies, VHH, or other domain structure), and functionally active epitope-binding fragments of any of the above. In particular, antibodies include immunoglobulin molecules and immunologically active fragments of immunoglobulin molecules, namely, molecules that contain an analyte-binding site. Immunoglobulin molecules can be of any type (for example, IgG, IgE, IgM, IgD, IgA, and IgY), class (for example, IgG1, IgG2, IgG3, IgG4, IgA1, and IgA2), or subclass. For simplicity’s sake, an antibody against an analyte is frequently referred to herein as being either an “anti-analyte antibody” or merely an “analyte antibody.”
JHU Ref. No. C18037_P18037-02 Atty. Docket No. JHU-42990.601 “Antibody fragment” as used herein refers to a portion of an intact antibody that retain the ability to specifically bind to an antigen (see, generally, Holliger et al., Nat. Biotech., 23(9): 1126- 1129 (2005)) (e.g., comprises the antigen-binding site or variable region). Any antigen-binding fragment of the antibody described herein is within the scope of the present disclosure. The antibody may not include the constant heavy chain domains (e.g., CH2, CH3, or CH4, depending on the antibody isotype) of the Fc region of the intact antibody. Examples of antibody fragments include, but are not limited to, Fab fragments, Fab’ fragments, Fab’-SH fragments, F(ab’)2 fragments, Fd fragments, Fv fragments, diabodies, single-chain Fv (scFv) molecules, single-chain polypeptides containing only one light chain variable domain, single-chain polypeptides containing the three CDRs of the light-chain variable domain, single-chain polypeptides containing only one heavy chain variable region, and single-chain polypeptides containing the three CDRs of the heavy chain variable region. Typically, an immunoglobulin or antibody is a protein that comprises at least one complementarity determining region (CDR). The CDRs form the “hypervariable region” of an antibody, which is responsible for antigen binding (discussed further below). A whole antibody typically consists of four polypeptides: two identical copies of a heavy (H) chain polypeptide and two identical copies of a light (L) chain polypeptide. Each of the heavy chains contains one N- terminal variable (VH) region and three C-terminal constant (CH1, CH2, and CH3) regions, and each light chain contains one N-terminal variable (VL) region and one C-terminal constant (CL) region. The light chains of antibodies can be assigned to one of two distinct types, either kappa (κ) or lambda (λ), based upon the amino acid sequences of their constant domains. In a typical antibody, each light chain is linked to a heavy chain by disulfide bonds, and the two heavy chains are linked to each other by disulfide bonds. The light chain variable region is aligned with the variable region of the heavy chain, and the light chain constant region is aligned with the first constant region of the heavy chain. The remaining constant regions of the heavy chains are aligned with each other. The variable regions of each pair of light and heavy chains form the antigen binding site of an antibody. The VH and VL regions have the same general structure, with each region comprising four framework (FW or FR) regions. The term “framework region,” as used herein, refers to the relatively conserved amino acid sequences within the variable region which are located between the CDRs. There are four framework regions in each variable domain, which are designated FR1, FR2, FR3, and FR4. The framework regions form the β sheets that provide the structural framework of the variable region (see, e.g., C. A. Janeway et al. (eds.), Immunobiology, 5th Ed., Garland Publishing, New York, N.Y. (2001)).
JHU Ref. No. C18037_P18037-02 Atty. Docket No. JHU-42990.601 As used herein, the term “close proximity” refers to two targets (e.g., target X and target Y) that are in physical or spatial proximity, either due to direct binding between the two targets or indirectly due to interactions between other molecules, cells, or the like. With regard to the methods disclosed herein, close proximity allows single detection of both targets with a single signal generating complex. In some embodiments, target X and target Y are on different molecules. For example, target X and target Y are different molecules and are in “close proximity” when they are present in the same complex, bound to the same binding partner (e.g., protein, nucleic acid, small molecule, drug), in a similar location (e.g., on a cell membrane or in the same organelle), or on two associated structures or cells. The term “contacting” as used herein refers to bring or put in contact, to be in or come into contact. The term “contact” as used herein refers to a state or condition of touching or of immediate or local proximity. The term “detecting” as used herein generally refers to any form of measurement, and includes determining whether an element is present or not. This term includes quantitative and/or qualitative determinations. The terms “nucleic acid,” “polynucleotide,” and “oligonucleotide” are used herein to describe a polymer composed of nucleotides, e.g., deoxyribonucleotides or ribonucleotides, or compounds produced synthetically, which can hybridize with naturally occurring nucleic acids in a sequence specific manner analogous to that of two naturally occurring nucleic acids, e.g., can participate in Watson-Crick base pairing interactions. As used herein in the context of a polynucleotide sequence, the term “bases” (or “base”) is synonymous with “nucleotides” (or “nucleotide”), the monomer subunit of a polynucleotide. The terms “nucleoside” and “nucleotide” are intended to include those moieties that contain not only the known purine and pyrimidine bases, but also other heterocyclic bases that have been modified. Such modifications include methylated purines or pyrimidines, acylated purines or pyrimidines, alkylated riboses or other heterocycles. In addition, the terms “nucleoside” and “nucleotide” include those moieties that contain not only conventional ribose and deoxyribose sugars, but other sugars as well. Modified nucleosides or nucleotides also include modifications on the sugar moiety, e.g., wherein one or more of the hydroxyl groups are replaced with halogen atoms or aliphatic groups, or are functionalized as ethers, amines, or the like. “Analogues” refer to molecules having structural features that are recognized in the literature as being mimetics, derivatives, having analogous structures, or other like terms, and include, for example, polynucleotides incorporating non-natural nucleotides, nucleotide mimetics such as 2’-modified nucleosides, peptide nucleic acids, oligomeric nucleoside phosphonates, and any polynucleotide that has added substituent groups, such as protecting groups or linking moieties.
JHU Ref. No. C18037_P18037-02 Atty. Docket No. JHU-42990.601 The term “complementary” refers to specific binding between polynucleotides based on the sequences of the polynucleotides. As used herein, a first polynucleotide and a second polynucleotide are complementary if they bind to each other in a hybridization assay under stringent conditions, e.g., if they produce a given or detectable level of signal in a hybridization assay. Portions of polynucleotides are complementary to each other if they follow conventional base- pairing rules, e.g., A pairs with T (or U) and G pairs with C, although regions (e.g., less than 5 nucleotides) of mismatch, insertion, or deleted sequence may be present. The term “protein interaction” as used herein refers to interactions between different proteins. The interactions may be due to direct binding (covalent or non-covalent) interaction or due to biochemical associations and processes which result in two different proteins to be in close proximity. For example, protein interactions include, but are not limited to, interactions between protein components of a single multi-protein complex, protein binding pairs, two proteins binding the same target, two proteins localizing to a single location within a cell, two proteins on two different cells being in close proximity due to cell-cell interactions, two parts, subunits, or domains of a protein being in close proximity or non-covalently interacting as a result of folding, unfolding, activation, post-translational processing, binding of a target ligand or substrate, and the like. As used herein, the term “sample” is used in its broadest sense. In one sense, it is meant to include a specimen obtained from any source, including biological samples. Biological samples may be obtained from animals (including humans) and encompass fluids, solids, tissues, and gases. Such examples are not however to be construed as limiting the sample types. Preferably, a sample is a fluid sample such as a liquid sample. Examples of liquid samples that may be assayed include bodily fluids (e.g., blood, serum, plasma, saliva, urine, ocular fluid, semen, sputum, sweat, tears, pleural effusions, ascites, thin needle aspirates and spinal fluid). Viscous liquid, semisolid, or solid specimens may be used to create liquid solutions, eluates, suspensions, or extracts that can be samples. For example, throat or genital swabs may be suspended in a liquid solution to make a sample. Samples can comprise biological materials, such as cells, microbes, organelles, and biochemical complexes. Liquid samples can be made from solid, semisolid, or highly viscous materials, such as fecal matter, tissues, organs, biological fluids, or other samples that are not fluid in nature. For example, solid or semisolid samples can be mixed with an appropriate solution, such as a buffer, a diluent, and/or extraction buffer. The sample can be macerated, frozen and thawed, or otherwise extracted to form a fluid sample. Residual particulates may be removed or reduced using conventional methods, such as filtration or centrifugation. “Biological sample,” “sample from a subject,” “test sample,” and “patient sample” as used interchangeably herein may be a sample of blood, such as whole blood (including for example,
JHU Ref. No. C18037_P18037-02 Atty. Docket No. JHU-42990.601 capillary blood, venous blood, dried blood spot, etc.), tissue, urine, serum, plasma, amniotic fluid, an anal sample (such as an anal swab specimen), lower respiratory specimens such as, but not limited to, sputum, endotracheal aspirate or bronchoalveolar lavage, nasal mucus, cerebrospinal fluid, placental cells or tissue, endothelial cells, leukocytes, or monocytes. The sample can be used directly as obtained from a patient or can be pre-treated, such as by filtration, distillation, extraction, concentration, centrifugation, inactivation of interfering components, addition of reagents, and the like, to modify the character of the sample in some manner as discussed herein or otherwise as is known in the art. A variety of cell types, tissue, or bodily fluid may be utilized to obtain a sample. Such cell types, tissues, and fluid may include sections of tissues such as biopsy and autopsy samples, disease samples (e.g., tumor samples), oropharyngeal specimens, nasopharyngeal specimens, nasal mucus specimens, frozen sections taken for histologic purposes, blood (such as whole blood, dried blood spots, etc.), plasma, serum, red blood cells, platelets, an anal sample (such as an anal swab specimen), interstitial fluid, cerebrospinal fluid, etc. Cell types and tissues may also include lymph fluid, cerebrospinal fluid, or any fluid collected by aspiration. A tissue or cell type may be provided by removing a sample of cells from a human and a non-human animal, but can also be accomplished by using previously isolated cells (e.g., isolated by another person, at another time, and/or for another purpose). Archival tissues or preserved tissues, such as those having treatment or outcome history, may also be used. A “subject” or “patient” may be human or non-human and may include, for example, animal strains or species used as “model systems” for research purposes, such a mouse models, prokaryotic models (e.g., bacteria), archea, and single-celled eukaryotes(e.g., yeast). Likewise, subject may include either adults or juveniles (e.g., children). Moreover, patient may mean any living organism, preferably a mammal (e.g., humans and non-humans) that may benefit from the uses of compositions and methods contemplated herein. Examples of mammals include, but are not limited to, any member of the Mammalian class: humans, non-human primates such as chimpanzees, and other apes and monkey species; farm animals such as cattle, horses, sheep, goats, swine; domestic animals such as rabbits, dogs, and cats; laboratory animals including rodents, such as rats, mice and guinea pigs, and the like. Examples of non-mammals include, but are not limited to, birds, fish, and the like. In one embodiment, the subject is a human. Preferred methods and materials are described below, although methods and materials similar or equivalent to those described herein can be used in practice or testing of the present disclosure. All publications, patent applications, patents and other references mentioned herein are
JHU Ref. No. C18037_P18037-02 Atty. Docket No. JHU-42990.601 incorporated by reference in their entirety. The materials, methods, and examples disclosed herein are illustrative only and not intended to be limiting. 2. Methods of Detection Protein Interactions in a Sample In one aspect, provided herein are methods for detecting protein interactions in a sample. The methods comprise contacting a sample with a first detection agent and a second detection agent, wherein each of the first and second detections agents comprise an antibody or fragment thereof directed to a first and second protein of interest, respectively, conjugated to a first and second barcode oligonucleotides, respectively. In some embodiments, the barcode oligonucleotides are configured to form a contiguous oligonucleotide when the first protein of interest and the second protein of interest are in close proximity. As such, the formation of the contiguous oligonucleotide facilitates detection and identification of a direct or indirect protein interaction between the first and second protein of interest. In some embodiments, the methods comprise: (i) contacting the sample with at least one first detection agent, wherein each of the at least one first detection agent comprises an antibody, or fragment thereof, to a first protein of interest covalently attached to a first barcode oligonucleotide; (ii) contacting the sample with at least one second detection agent, wherein the second detection agent comprises an antibody, or fragment thereof, to a second protein of interest covalently attached to a second barcode oligonucleotide, wherein the first barcode oligonucleotide and the second barcode oligonucleotide are configured for ligation to each other when in close proximity; (iii) ligating the first barcode oligonucleotide and the second barcode oligonucleotide to form a contiguous oligonucleotide; and (iv) detecting formation of the contiguous oligonucleotide. In some embodiments, the methods comprise: (i) contacting the sample with at least one first detection agent, wherein each of the at least one first detection agent comprises an antibody, or fragment thereof, to a first protein of interest covalently attached to a first barcode oligonucleotide comprising a first cleavage site; (ii) contacting the sample with at least one second detection agent, wherein the second detection agent comprises an antibody, or fragment thereof, to a second protein of interest covalently attached to a second barcode oligonucleotide comprising a second cleavage site; (iii) contacting the sample with one or more cleavage agents configured to act on the first cleavage site and second cleavage site and a ligase; and (iv) detecting formation of a contiguous oligonucleotide comprising the first barcode oligonucleotide and the second barcode oligonucleotide. The first and second detection agents utilize antibodies or a fragment thereof, which are covalently attached to a barcode oligonucleotide. The antibodies or the fragment thereof bind to a protein of interest in the sample, and their respective barcode oligonucleotide provide a sequence
JHU Ref. No. C18037_P18037-02 Atty. Docket No. JHU-42990.601 which can be used to identify the protein of interest but also provide a cleavage sequence which allows formation a single contiguous oligonucleotide when the first and second proteins of interest are in close proximity. The methods disclosed can be used to target a single first or second protein of interest. The methods of detecting protein interactions disclosed herein can be used for concurrent or sequential detection of multiple protein interactions in the same sample. For example, in some embodiments, the method comprises detecting two or more protein interactions in the same sample. In some embodiments, the method comprises detecting 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, or more different protein interactions in the same sample. For example, in some embodiments, the method comprises detecting from 1 to 100 different protein interactions in the sample. In some embodiments, the method comprises detecting from 1 to 50 different protein interactions in the sample. In some embodiments, the method comprises simultaneously or sequentially contacting the sample with a plurality of first detection agents, each directed to a different first protein of interest, and/or contacting the sample with a plurality of second detection agents, each directed to a different second protein of interest. Thus, the methods can be multiplexed to target a vast array of protein interactions in a single sample. The methods disclosed can be used to study protein interactions in bulk samples or single cell samples. Any suitable antibody or fragment thereof can be used in the first and second detection agents. Preferably, the antibodies or fragments thereof are validated with high specificity and selectivity. In some embodiments, the antibody or a fragment thereof binds directly to the protein of interest. Alternatively, the antibody may bind indirectly to the protein of interest (e.g., to a primary antibody bound to the protein of interest). In some embodiments, the barcode oligonucleotides are fully or partially double stranded. The barcode oligonucleotides comprise a barcode and a unique molecular identifier (UMI) flanked by primer binding sites. Each of the sequences for the barcode, UMI, and the primer binding sites can be separated by spacer base pairs or can be immediately adjacent to each other. In some embodiments, the barcode oligonucleotides each comprise a cleavage site. In some embodiments, the cleavage sites are distal to the barcode and the UMI in relationship to the site of attachment the antibody. Thus, when the barcode oligonucleotides are cleaved, the barcode and UMI remain conjugated to the antibody of interest. Accordingly, after cleavage if the generated single stranded ends on the oligonucleotide hybridize with another barcode oligonucleotide, due to being close in proximity, the barcode and UMI are incorporated into the contiguous oligonucleotide.
JHU Ref. No. C18037_P18037-02 Atty. Docket No. JHU-42990.601 Any cleavage sites which create single stranded ends for use in binding to complementary single stranded oligos can be used in the methods. In some embodiments, the cleavage sites in the first detection agent or second detection agent are the same. In some embodiments, the cleavage sites in the first detection agent or second detection agent are different. In either instance, the single stranded ends created are configured to be complementary to hybridize and form a single contiguous oligonucleotide when treated with a ligase when in close proximity. The barcode oligonucleotides comprise a barcode sequence. The barcode sequence may be any length or sequence. The barcode sequence can be used to specify if the detection agent is a first detection agent or a second detection agent. For example, the barcode sequence can be used to specify if the detection agent is bound or configured to bind to a solid surface, as described elsewhere herein. In some embodiments, the barcode oligonucleotides comprise a unique molecular identifier (UMI) sequence. The UMI may be any suitable sequence of nucleic acids of any suitable length. The UMI may be a sequence specifically correlated with a specific protein of interest, such that identifying the UMI allows identification of the protein of interest to which the first or second detection agent is targeted. UMIs can also be used to account for PCR and sequencing artifacts in subsequent sequencing analysis. The barcode oligonucleotides can vary in length based on the size of the described components. For example, the barcode oligonucleotides can have a length of about 20 to about 150 basepairs. The barcode oligonucleotide is covalently attached to the antibody or fragment thereof. In some embodiments, the covalent attachment is via a direct bond between the antibody or the fragment thereof and the barcode oligonucleotide. In some embodiments, the barcode oligonucleotide is covalently attached via a linker. General methods of conjugating oligonucleotides to antibodies are known to those skilled in the art. For example, a typical conjugation method includes use of a linker compound that includes two distinct reactive moieties, which react with different types of functional groups (e.g., one group that reacts with an amine, such as an activated ester group, and one group that reacts with a thiol, such as a maleimide group). Such reactive moieties used in conjugation reactions are well-known to those skilled in the art, and include activated esters such as succinimidyl and sulfosuccinimidyl esters and pentafluorophenyl esters, maleimides, azides, alkynes, hydrazines, isocyanates, isothiocyanates, haloacetamides, and the like. Methods of installing such reactive groups are well-known to those skilled in the art. As one non-limiting example, an amino group can be installed at the 5’-end of an oligonucleotide via phosphoramidite chemistry.
JHU Ref. No. C18037_P18037-02 Atty. Docket No. JHU-42990.601 In some embodiments, the linker can include one or more nucleotides. For example, the linker can comprise an oligonucleotide sequence. Such a sequence may be considered separate from the oligonucleotide sequence to which the nucleic acid component of the signal-generating complex can hybridize. The linker can include additional atoms or groups; for example, if the antibody is reacted with 2-iminothiolane, it is understood that the linker will further include atoms derived from such reaction. Accordingly, in some embodiments, the linker further comprises one or more additional groups selected from -CH2-, -O-, -NH-, -S-, -C(=O)-, -C(=NH)-, and any combination thereof (e.g., combinations of such moieties could include ester groups (-C(=O)O-), amide groups (-C(=O)NH-), carbamate groups (-NHC(=O)O-), ethylene glycol groups (-CH2CH2O-), and the like. In some embodiments, the linker comprises an antibody-binding domain. In some embodiments, an antibody binding domain (AbBD) comprises Protein A, Protein G, Protein L, CD4, or a fragment thereof. In some embodiments, the antibody-binding domain is an engineered antibody-binding domain, such as to include a non-natural amino acid, a photoreactive group, or a crosslinker. In some embodiments, the antibody binding domain is operably linked to a photoreactive amino acid group, for example, benzoylphenylalanine (BPA), resulting in a photoreactive antibody binding domain (pAbBD). In some embodiments, the antibody-binding domain (AbBD) is operably linked to a photoreactive amino acid which is operably linked to an antibody or a fragment thereof. In some embodiments, an antibody is first reacted with the linker compound to provide a functionalized antibody, which is subsequently reacted with the barcode oligonucleotide to provide the detection agent. In other embodiments, a barcode oligonucleotide is first reacted with the linker compound to provide a functionalized barcode oligonucleotide, which is subsequently reacted with an antibody to provide the detection agent. In some embodiments, an antibody is first conjugated with anchor oligonucleotide. The anchor oligonucleotide comprises a nucleic acid sequence which hybridizes to at least a portion of a single-stranded barcode oligonucleotide. To form a double-stranded barcode oligonucleotide the anchor oligonucleotide can be extended using known DNA synthetic enzymes and methods. The method is not limited by the order in which steps (i) and (ii) are performed. In some embodiments, step (i) and step (ii) are performed simultaneously. In some embodiments, step (i) is performed before step (ii). In some embodiments, step (ii) is performed before step (i). In some embodiments, step (i) and/or step (ii) comprises contacting the sample with the first and/or second detection agent for about 10 minutes to about 48 hours. In some embodiments, step (i) and/or step (ii) comprises contacting the sample with the first and/or second detection agent at a temperature of about 4 °C to about 25 °C.
JHU Ref. No. C18037_P18037-02 Atty. Docket No. JHU-42990.601 In some embodiments, the methods further comprise separating or isolating detection agents and binding partners from the sample. Separating or isolating the detection agents and their respective binding partners from the remainder of the sample facilitates analysis of those protein interactions formed in step (i) and step (ii). In some embodiments, the first detection agent or second detection agent is bound to a solid surface or comprises a functional group configured to bind to a solid surface. When bound to the solid surface the remainder of the sample can be washed away or merely removed from that portion of the sample tethered or associated to the solid surface. The solid surface may be a bead or other particle, a microplate, a microfluidic device, and the like. In some embodiments, detecting the contiguous oligonucleotide comprises isolating contiguous oligonucleotides from the sample and amplifying and/or sequencing the contiguous oligonucleotides. Any methods known in the art for purifying or separating nucleic acid can be used to isolate the contiguous oligonucleotides. In some embodiments, the method comprises treating the sample or isolated detection agents and binding partners with a protease followed by DNA extraction. In some embodiments, the protease is selected from trypsin, proteinase K, pepsin, pronase, endoproteinase AspN, and endoproteinase GluC. Any suitable amplification method known in the art allowing for sensitive detection of DNA may be used, including by not limited to polymerase chain reaction (PCR), preferably real time PCR. Sequencing can be accomplished using high-throughput systems, some of which allow detection of a sequenced nucleotide immediately after or upon its incorporation into a growing strand, e.g., detection of sequence in real time or substantially real time. In some embodiments, or sequencing the contiguous oligonucleotides is achieved by next-generation sequencing. In some embodiments, the methods further comprise mapping the contiguous oligonucleotides to the protein interactions based on the barcode and/or unique molecular identifier (UMI) sequences. Mapping the contiguous oligonucleotides can comprise extracting the barcodes, primer sequences, UMI, and/or linkers from the detection agents, sequencing the contiguous oligonucleotides, and comparing them to the sequence files created during the analysis to identify those sequences which were found in contiguous oligonucleotides. Further the sequences can be assigned to protein pairs through alignment. Once the pairs are generated, the methods may further comprise counting protein pairs, generating statistics for individual proteins, protein pairs, types of proteins. This data can be used to examine and determine abundance of the individual proteins, protein pairs, types of proteins. Details for the mapping and subsequent analysis is provided in Examples 1, 2, and 6 below. Mapping and subsequent analysis may be completed using computer implemented methods, also provided in the present disclosure.
JHU Ref. No. C18037_P18037-02 Atty. Docket No. JHU-42990.601 In some embodiments, the sample is a biological sample. The biological sample can be derived from various sources. In some embodiments, the biological sample is a tissue specimen or is derived from a tissue specimen. In some embodiments, the biological sample is a blood sample or is derived from a blood sample. In some embodiments, the biological sample is a cytological sample or is derived from a cytological sample. In some embodiments, biological sample is cultured cells. Any manner of protein interactions can be detected by the methods disclosed herein. The methods may detect interactions between two different proteins. For example, in some embodiments, the first detection agent binds, directly or indirectly, to a first protein of interest, and the second detection agent binds, directly or indirectly, to a second protein of interest. Close proximity may indicate, for example, that the two different proteins localize to similar structures in the same cell, are within the same multi-protein complex, associate with a common binding partner, or are direct binding partners to each other. The methods are not limited by the proteins of interest. The proteins of interest may include cell cycle proteins, signal-pathway related proteins, disease-related proteins, transcription factors, and the like. The proteins of interest may be cancer-related proteins. For example, the proteins of interest may include transcription factors with known roles in cancer, e.g., BACH1/2, BRD9, CEBP family members, CREB5, ERG, FOS/JUN and related proteins, GATA4, FOX family members, KLFs, MYC and MYC-associated factors, SMAD family members, SOX family members, STAT family members, and multiple zinc fingers, as well as HDACs and other chromatin-associated proteins. 3. Systems Disclosed herein are systems for detecting protein interactions. In some embodiments, the systems comprise at least one first detection agent, wherein each of the at least one first detection agent comprises an antibody, or fragment thereof, to a first protein of interest covalently attached to a first barcode oligonucleotide; and at least one second detection agent, wherein the second detection agent comprises an antibody, or fragment thereof, to a second protein of interest covalently attached to a second barcode oligonucleotide. In some embodiments, the first barcode oligonucleotide and the second barcode oligonucleotide are configured for ligation to each other when in close proximity. In some embodiments, the first barcode oligonucleotide and the second barcode oligonucleotide are fully or partially double stranded. In some embodiments, the first barcode oligonucleotide and the second barcode oligonucleotide each contain a cleavage site configured to create complementary single strand ends.
JHU Ref. No. C18037_P18037-02 Atty. Docket No. JHU-42990.601 The barcode oligonucleotides comprise a barcode and a unique molecular identifier (UMI) flanked by primer binding sites. Each of the sequences for the barcode, UMI, and the primer binding sites can be separated by spacer base pairs or can be immediately adjacent to each other. Descriptions provided above to the barcode, UMI and cleavage sites for the disclosed methods are also applicable to the systems disclosed herein. In some embodiments, one of the first detection agent or second detection agent is bound to a solid surface or comprises a functional group configured to bind to a solid surface. In some embodiments, the solid surface is a bead or particle. The solid surface may be provided separately from the first detection agent or second detection agent, and accordingly, the system may comprise reagents to conjugate the first detection agent or second detection agent to the solid surface. The systems disclosed herein can be used to detect any number of protein interactions simultaneously or sequentially. In some embodiments, the system comprises a plurality of first detection agents, each directed to a different first protein of interest. In some embodiments, the system comprises a plurality of second detection agents, each directed to a different second protein of interest. In some embodiments, the system further comprises one or more primers configured to bind to the primer binding sites on the first barcode oligonucleotide and/or the second barcode oligonucleotide. The systems may further comprise one or more reagents necessary for protease digestion, DNA extraction, nucleic acid amplification (e.g., PCR), and nucleic acid sequencing. Many such reagents are known in the art and commercially available. Examples of suitable reagents include conventional reagents employed in nucleic acid amplification reactions, such as, for example, one or more enzymes having polymerase activity, enzyme cofactors (such as magnesium or nicotinamide adenine dinucleotide (NAD)), salts, buffers, deoxyribonucleotide, or ribonucleotide triphosphates (dNTPs/rNTPs; for example, deoxyadenosine triphosphate, deoxyguanosine triphosphate, deoxycytidine triphosphate, and deoxythymidine triphosphate) blocking agents, labeling agents, and the like. Other reagents used in amplification reactions include nicking enzymes, single-strand binding proteins, helicases, resolvases, and the like. The systems may also further comprise equipment to perform any of the protease digestion, DNA extraction, nucleic acid amplification (e.g., PCR), and nucleic acid sequencing, including but not limited to incubators, centrifuges, thermocyclers, sequencers, and the like.
JHU Ref. No. C18037_P18037-02 Atty. Docket No. JHU-42990.601 4. Kits In another aspect, the disclosure provides kits comprising at least one first detection agent and/or and at least one second detection agent, as described herein. In some embodiments, the kits comprise one or more components to generate at least one first detection agent and/or and at least one second detection agent, as described herein. For example, the kit may include one or more antibodies, functionalized with a linker or an anchor oligonucleotides, a single stranded barcode oligonucleotide, functionalization reagents, DNA synthesis reagents, and the like. The kits can also comprise other components necessary for carrying out the disclosed methods, including primers, cleavage agents, proteases, ligases, DNA purification reagents, as described elsewhere herein. The kits can also comprise instructions for using the components of the kit. The instructions are relevant materials or methodologies pertaining to the kit. The materials may include any combination of the following: background information, list of components, brief or detailed protocols for using the compositions, trouble-shooting, references, technical support, and any other related documents. Instructions can be supplied with the kit or as a separate member component, either as a paper form or an electronic form which may be supplied on computer readable memory device or downloaded from an internet website, or as recorded presentation. It is understood that the disclosed kits can be employed in connection with the disclosed methods. The kit may further contain containers or devices for use with the methods or compositions disclosed herein. The kits optionally may provide additional components such as buffers and disposable single-use equipment (e.g., pipettes, cell culture plates, flasks). The kits provided herein are in suitable packaging. Suitable packaging includes, but is not limited to, vials, bottles, jars, flexible packaging, and the like. Individual member components of the kits may be physically packaged together or separately. 5. Examples Example 1 Label-free analysis of protein-protein interactions A resource of stringently validated monoclonal mAbs targeting human transcription factors has been created and shared (see Example 2 for additional details). These validated mAbs can be used to detect interactions between unlabeled proteins in native cells (Fig.2). mAbs for hundreds of proteins of interest are used to create two modified sets: Set A, which is biotinylated and barcoded with a specific dsDNA sequence, and Set B, which is barcoded with a different dsDNA sequence but not biotinylated. The Set A barcodes share a single anchor sequence (Fig.2, red line), and the Set B
JHU Ref. No. C18037_P18037-02 Atty. Docket No. JHU-42990.601 barcodes share a different anchor sequence (Fig.2, blue line). The Set A mAbs are separately immobilized on streptavidin beads, washed stringently, and pooled, and the Set B mAbs are also pooled. Cell populations of interest are then lysed, incubated overnight at 4° C with aliquots of the Set A and Set B pools, and subjected to three high-salt washes and three low-salt washes. When a protein is captured by a Set A mAb immobilized on the beads, and its binding partner is recognized by Set B mAb, the two DNA barcode sequences are brought into proximity. Golden Gate Assembly (BsaI and T4 DNA ligase) reactions ligate the two barcode sequences into a single piece of DNA. After Proteinase K digestion, the ligated products can be analyzed. The Golden Gate method permits only Set A – Set B barcode ligation; neither A-A nor B-B barcodes are observed. Furthermore, directional sequencing yields all forward-strand sequences in the A-B orientation, with reverse complements not observed. Example 2 Detect protein-protein interactions with label-free proteins and barcode sequencing The quality, consistency, and availability of research-grade antibodies present continuing problems. Global estimates are over $800 million wasted annually by using poor-quality antibodies. Aggravating factors include the absence of standardized antibody-validation criteria in the research community, a lack of transparency from commercial antibody suppliers about their products, the use of polyclonal reagents with extensive batch variation, and technical difficulties in comprehensive assessments of antibody cross-reactivity. In 2010, the NIH initiated the Protein Capture Reagents Program (PCRP) to address these challenges. As one of the two Production Centers, a collection of 1,406 highly specific and IP- and immunoblot (IB)-validated mouse monoclonal antibodies (mAbs) targeting 736 unique human TFs was generated (Fig.3). HuProt arrays were first used, each comprising 20,240 individual purified human proteins, to identify mAbs that recognize their cognate targets with high specificity. Using an integrated pipeline, of the 1406 mAbs, 808, 544, and 54 were validated as IP-grade, IP- and IB-grade, and IB-grade, respectively. Some mAbs were shown to be ChIP-seq- or IHC-grade. This collection of mAbs has become a unique set of high-quality affinity reagents to characterize TFs in vivo. MuDPPI Barcoding The procedure of barcoding mAbs proceeded with dsDNA via covalent conjugation (Fig.4). Briefly, purified mAbs were individually conjugated to a thiolated “anchor” oligo (25-mer) via a bifunctional linker, SMCC. Next, a collection of DNA barcode oligos, each comprising a 5’-Primer 1 (25-mer) with a BsaI restriction site, an 11-mer DNA barcode, an 8- mer UMI, and a 3’-Primer 2 (25-mer, complementary to the anchor oligo), were synthesized (Fig. 4A). Anchor oligo-tethered mAbs were then individually annealed to a particular DNA barcode oligo and converted to dsDNA. As judged by up-shifted bands as compared with un-treated mAb (Fig.4B,
JHU Ref. No. C18037_P18037-02 Atty. Docket No. JHU-42990.601 lane 1), all mAbs were successfully conjugated to the DNA barcodes. The success of synthesis of ds- DNA barcodes conjugated to mAbs was also judged by up-shifting of mAb bands (red arrows) and labeling with Cy5-labeled dCTP after the Klenow reactions (Fig.4C). RT-PCR analysis of targeted pairs HeLa cells were used to examine heterodimer formation of several well-studied human TF pairs. Approximately 10,000 HeLa cells were grown to confluence, lightly fixed (0.4% formaldehyde for 10 min), and lysed. An aliquot of 5 pooled Set A barcoded mAbs (anti-Fos, -Jun, -Max, and -Mad1, with anti-GPDH as control) on beads (~50 μL bead slurry) was added to the cell lysate, followed by adding an aliquot of the corresponding Set B mAbs. After overnight incubation at 4° C, the beads were subjected to three high-salt washes (500 mM NaCl in TE buffer at pH 7) and three low-salt washes (100 mM NaCl in TE) at RT. After resuspending the beads in T4 DNA ligase buffer, Golden Gate Assembly Mix (New England Biolabs) was added, followed by several cycles of incubation between 37° C and 14° C. Next, Proteinase K was added to the beads to release the ligated DNA products. After heat-inactivation, the supernatants were collected and subjected to real-time PCR reactions using specific primer pairs. With GAPDH as a negative control, -ΔCt values were obtained by subtracting the GAPDH Ct value from each corresponding TF pair’s Ct value. As expected, the -ΔCt values of the well-established heterodimers of Max-Mad1 and Jun-Fos were significantly higher (P < 0.001) than the GPDH negative controls, no matter which antibody of each pair was immobilized on streptavidin beads (Fig. 5). We also observed the formation of Max, Jun, Fos, and Mad1 homodimers as reported previously. NextGen sequencing Rather than the RT-PCR readout, PCR amplification of the ligated barcodes can precede Illumina sequencing, with 3 independent replicates for each lysate. As described above, the total barcode length is 88 bp (20nt primer + 11nt barcode + 8nt UMI for Set A and Set B, plus 10nt total linker sequence), and >100nt paired-end reads are used for overlapping coverage of the barcode. FLASH merges the read-pairs prior to paring the barcodes, reducing to UMIs if desired, and counting reads or UMI events for each protein pair and orientation, e.g., with computer implemented software or methods. Standard sequence quality controls ensure that all barcodes are of the proper A-B form expected from Golden Gate Assembly and assess the ability to map reads to barcodes without error. Multiplexing Sequencing data using HeLa cells with a library of 178 proteins in Set A and Set B (15.9K possible heterodimers and homodimers) was generated.2.4M Illumina reads were produced, of which 1.8M were mapped to an expected barcode ligation product (75% mapping rate). Of the 178 proteins, which included GAPDH as a negative control, >100 reads were identified for 167 proteins (94%).30 protein interactions with barcodes enriched at >2-fold, 169 interactions enriched >1.5-fold, and 7822 interactions with 100+ total reads were identified. A second assay used
JHU Ref. No. C18037_P18037-02 Atty. Docket No. JHU-42990.601 a mixture of 154 mAbs against 10,000 HEK293 cells. After a total of 20 PCR cycles, the amplicons were running at the expected size range (229-243 bp), and 4M reads were obtained with i-Seq. Interaction detection Statistical analyses identify interactions whose barcode counts are significantly over-represented accounting for multiple testing of all 15K+ possible pairs. For simplicity, “read count” is used to refer to raw count or to UMI-reduced events. Considering a specific pair of proteins denoted 1 and 2, let n12 denote the read count for the configuration with protein 1 with a Set A barcode and protein 2 with a set B barcode, and n21 denote the swapped orientation with protein 2 from Set A and protein 1 from Set B. Let a1 and a2 represent the number of reads with protein 1 or protein 2 from Set A, and b1 and b2 the corresponding counts for proteins 1 or 2 from Set B. Denoting T as the total number of reads mapping to barcodes, and assuming a simple null hypothesis of random pairing, n12 is approximately Poisson-distributed with an expected value of a1b /T. Similarly, the sum of counts in both orientations, n12+n21 under the null is a Poisson- distributed random variable with expected value (a1b2+a2b1)/T. These statistics are used to calculate a one-sided single-test p-value for each pair to have counts higher than expected under the null. We will also calculate a nominal log2-enrichment as the log2 of the ratio of observed to expected counts, using standard add-1 pseudocounts to avoid numerical errors for pairs with zero observed counts. Volcano plots of –log10(p-value) vs log2- enrichment to visualize results and set a 0.05 family-wise error rate of ~0.05/15K = 3x10–6 for significance. Ability to detect 2-fold enrichment is typical for power calculations with sequencing readout. For large counts, Poisson distributions are approximately normal. A p-value of 3x10–6 corresponds to a z-score of –4.5, and for 50% power to detect 2-fold enrichment the expected count would be 4.52 = 20. Summed over all 15K pairs suggests 300K read pairs. Validation A statistical validation comparing interactions detected by MuDPPI with interactions measured experimentally as curated by the Human Reference Interactome (HuRI) database can calculate putative false-positive and false-negative rates. Interaction predictions are also correlated with functional associations based on data integration (features include gene co- occurrence, gene co-expression, pathway membership, and text mining) available from the String database. Functional association scores facilitate selection of at least 6 newly discovered TF heterodimers for validation with orthogonal methods, such as IP-WB and IP-MS/MS, and an additional 6 known interactions also observed by MuDPPI as positive controls. Negative controls use anti-GAPDH and anti-IgG2b. A human ORF collection, comprising 21,000+ full- length human ORFs, can be used to subclone them to a mammalian expression vector that will express TF proteins C-terminally tagged with 3xFLAG. To test a given TF pair, one protein will be transiently over expressed in HeLa cells, immunoprecipitated with anti-FLAG(pCSF107mT-GATEWAY-3'-FLAG;
JHU Ref. No. C18037_P18037-02 Atty. Docket No. JHU-42990.601 AddGene) and western-blotted with an antibody targeting the other TF protein. The identity of co- immunoprecipitated proteins of each bait TF will be determined using a tandem mass tag (TMT) MS/MS approach. Example 3 Differential interaction analysis The mitotic cell cycle system in HEK293 cells was interrogated to determine whether MuDPPI technology can be used to monitor changes in PPIs, (Fig.6). MuDPPI was used to profile the abundance of CDK6/Cyclin D1 (peak at G1) and CDK1/Cyclin B (peak at M). The cell cycles of HEK293 cells were synchronized by starving the cells in serum-free media for 24 hours, followed by releasing the serum starvation with serum at different time points so that cells entering M, G2, S and G1 phases were collected at the end. Using the MuDPPI approach, CDK-cyclin interactions were quantified using RT-PCR with primer pairs specific for CDK6/Cyclin D1 and CDK1/Cyclin B barcodes. The Ct values of GAPDH-cyclin pairs were subtracted from those of each CDK/cyclin pair to obtain -ΔCt values. As expected, the CDK6/Cyclin D1 interaction was most abundant at G1 phase, and CDK1/cyclin B peaked at M phase, demonstrating that MuDPPI can quantitatively measure PPIs in cells. Detection of phosphorylation impact on protein-protein interactions. Interactions between two proteins can happen very quickly upon external stimulation and are often regulated by protein posttranslational modifications (PTMs). EGFR signaling in HEK293 cells was used to demonstrate that MuDPPI technology can detect transient changes in signal transduction networks arising from PTM-regulated protein-protein interactions (Fig.7). After HEK293 cells were serum-starved for 20 hr, serum and EGF (100 ng/mL) were introduced. Cells were collected at 0-, 2-, 4-, 10-, and 13-min post-treatment and immediately rinsed in ice-cold PBS. Each sample was then fixed under mild conditions (0.4% formaldehyde for 10 min) and lysed. MuDPPI assays were then carried out using mAbs targeting the whole protein and specific phosphorylation sites. The ΔCt values were obtained by subtracting Ras-RafpSer259 Ct values from the Ras-Raf Ct values at each time point post EGF treatment. The ΔCt values, equal to log2(Ras-Raf/Ras-RafpSer259), were significantly higher after 4 min versus time 0 (Fig.7). These results suggest that fully activated pRaf is more likely to dissociate from GTP-bound Ras, as reported in the literature, and demonstrate that MuDPPI has power to detect the regulation of PPIs by phosphorylation and other PTMs. To detect differences in protein interactions relevant to cell state, differential interaction analysis can be performed on two widely used cell line models for breast cancer, MCF7 (ATCC HTB-22), a weakly invasive model, and MDA-MB-231 (ATCC CRM-HTB-26), a highly invasive model, both derived from breast adenocarcinomas of epithelial origin. For differential interaction
JHU Ref. No. C18037_P18037-02 Atty. Docket No. JHU-42990.601 analysis, barcode counts (whether read counts or UMI counts) will be used as input to deseq2 for difference finding. A 0.05 family-wise error rate will be used to identify interactions that are differentially represented in the low-invasive vs high-invasive cell lines. Example 4 Detecting interactions in single cells The detection limit of MuDPPI was explored using sorted HEK293 cells (Fig.8). Flow cytometry was used to deposit 1 and 50 HEK293 cells into two groups of 16 wells, and the formation of three known protein pairs: Jun-JunB, Jun-Fos, and JunB homodimer, was examined with GAPDH again serving as the negative control. The averages of 16 Ct values of each expected pair are significantly lower (meaning stronger binding) than the corresponding GAPDH controls at the single-cell level (Fig.8). The results obtained with 50-cell pools were very similar to the single cells, suggesting that MuDPPI can detect PPIs at the single-cell level. Example 5 MuDPPI analysis of cell cycle in HEK 293T samples down to single cells To explore protein interactions across different cell-cycle phases, HEK 293T cell line cells were sorted using flow cytometry into 384-well plates at varying cell numbers, ranging from 1 to 250 cells per well. MuDPPI was used to detect protein interactions within these cell quantities (Table 1). Cell cycle synchronization was used to generate samples at different cell cycle phases. Initially, HEK cells were split and allowed to incubate for 2 days in serum-containing media to promote growth and proliferation. Subsequently, the cells underwent serum starvation by switching to serum-free media, inducing a quiescent state for cell cycle synchronization. After serum starvation, cells were selectively released into different cell-cycle phases by reintroducing serum- containing media at specific time points. Key measures of the cell cycle include ploidy and cell size. In the G1 phase, cells have 2N ploidy and prepare for DNA replication. During the S phase, DNA is duplicated to 4N ploidy in preparation for cell division. The final phase, M phase, involves synthesis of proteins and organelles to support mitosis and division into two 2N daughter cells. Additionally, there is a quiescent phase, G0, where cells are not actively dividing. Accurate sorting of cells based on their cell-cycle phases was achieved through gentle dissociation using TrypLE, followed by washing in HBSS, filtering the sample, and seeding it into 384-well V-bottom plates at varying densities. The staining of cells with a cell cycle reagent allows DNA to bind the dye stoichiometrically, enabling the determination of DNA content. The flow cytometric analysis of cell count versus linear fluorescence generates a histogram of DNA content distribution for the cell cycle. A standard modeling algorithm can be used to determine whether the cells are in the G0/G1 phase, S phase, G2, or polyploidy state.
JHU Ref. No. C18037_P18037-02 Atty. Docket No. JHU-42990.601 To elucidate protein interactions within cell-cycle phases, 163 monoclonal antibodies (mAbs) recognizing 162 distinct proteins (2 of the mAbs were anti-FOS) including transcription factors, Cyclins A/B/D1/E, and CDKs 1/2/4/6, were employed to allow probing of the dynamic landscape of protein interactions with high specificity and sensitivity. Table 1 provides a summary of the samples. In all, 720 samples containing a total of 64,992 cells were analyzed. Table 1 - A total of 720 samples were collected according to cell cycle phase (G1 or S) with 1, 50, 100, or 250 cells included in the sample. Cell cycle phase Number of cells in sample Number of sample replicates G1 1 96
e average num er o rea -pa rs per samp e was , w . m on o a rea s. On average 66% of reads matched an expected A-B barcode without error, yielding 3.2 million total read-pairs, or 4470 per sample. Analysis based on individual protein counts The number of UMIs for each protein in each sample were counted, adding together the number of times the protein appeared on the A-side or B- side of an interaction. The data set was filtered to include samples with at least 1000 counts (676 of 720 samples) and proteins that appeared in at least 10 samples (162 of 163 proteins). The counts were normalized to 10,000 counts per cell, applied PCA, and plotted the samples along PC1 and PC2. Samples were grouped according to cell cycle phase, G1 vs. S, including samples with only a single cell (FIG.9A). These results demonstrated the ability of MuDPPI to generate high-quality data from single cells. Analysis based on protein pair interaction counts The same data was analyzed using pair interaction counts. Again, cells clustered by cell cycle phase (FIG.9B). A second axis showed variation with the number of cells in the sample. Example 6 MuDPPI computational methods Parsing MuDPPI sequencing library design. The MuDPPI sequencing library is provided as a table of mAbs, usually identified by gene symbol, and nucleotide sequences for the A-side and
JHU Ref. No. C18037_P18037-02 Atty. Docket No. JHU-42990.601 B-side ligation products. These sequences are parsed to extract the A-side and B-side barcodes and the expected locations of the primer sequences, the UMIs, and the linker regions of barcode-pair ligation products. Creating sample table. Given a directory with fastq or compressed fastq files, the sample identifiers are gathered and files that contain paired reads for the same sample are matched. Filing naming conventions can distinguish control vs. experimental samples, experimental groups such as cell type or cell cycle phase, or number of cells and that information is parsed as well. A sample table containing sample names, sample information, and locations of read files is generated based on the aforementioned. Merging reads. If read pairs are overlapping, the reads are merged into a single sequence for improved signal quality and simplified downstream analysis. The software FLASH is used for read merging.97-99% of the read pairs are typically able to be merged. Creating Zhu alignment/map (ZAM) file for sequence reads. The merged reads are analyzed to identify the A-side and B-side barcodes and assign each read to an A-B pair. Also the UMI sequence are provided for UMI-based read counting. Reads with mismatches are discarded. Perfect matching is used for speed. This approach typically yields perfect matches for 60-70% of reads. Imperfect matches could be implemented as an alternative. A pseudo-transcriptome of all possible barcode ligation products could be generated, and then an aligner-mapper such as bowtie2, hisat, or bwa, or a mapper such as salmon could be used to assign reads with mismatches permitted. Counting AB pairs. The ZAM file is parsed and for each possible AB pair, the total number of reads (‘total’ mode) and the number of unique UMIs among the total (‘UMI’ mode) is provided. Only reads with both A-side and B-side matches are counted. Generating statistics for proteins. The AB pair counts are used to generate statistics for each protein in the MuDPPI library. Statistics include the number of reads for the protein on the A- side, the number of reads on the B-side, and the sum of A-side and B-side, both for the total mode and the UMI mode of analysis. mAbs with an imbalance, with reads observed on one side but not the other, could be identified based on the statistics. Generating statistics for protein pairs. Protein-level statistics and the total number of observations are used to calculate a null hypothesis for the expected number of observations, a log- ratio of observed to expected observations (with and add-1 estimator to avoid undefined values), and a p-value for enrichment or depletion relative to the null. The null expectation is calculated according to a model in which A-B pairs are randomized, keeping the expected number of A-side and B-side observations for each protein equal to the observed numbers from the protein statistics. The p-values are modeled as a Poisson distribution, although other approaches such as a negative bionomial
JHU Ref. No. C18037_P18037-02 Atty. Docket No. JHU-42990.601 distribution with dispersion are possible. Statistics are generated for AB and BA orientations, the combined number of AB and BA orientations, and for total mode and UMI mode analysis. Generating group statistics. The sample table generated above is used to define sets of samples that should be grouped for analysis together. For example, given samples that are single cells, combine the single cell read counts together into an aggregated data set. This is implemented by first calculating AB counts for each sample individually as above. Then, the data sets from these files are streamed through the methods to calculate protein and protein-pair statistics. Analyzing differential abundance of proteins and interactions. One approach to analyze differential abundance of interactions is to use group statistics to identify protein-pair interactions that are significantly present in one sample or sample group but not in a second sample or sample group. This can be accomplished directly from the statistics generated for each sample or group. An alternative approach is to use count-based comparisons implemented by software for RNA sequencing analysis, for example deseq2 or edgeR. Protein comparisons are possible with standard normalization approaches. For protein-pair interaction analysis, raw counts may be used to assess the overall abundance of an interacting pair. The raw approach tests the hypothesis that the overall level of a protein complex has changed. An alternative is to normalize the raw counts by the expected number of observations. The normalized approach tests the hypothesis that protein complex membership has changed. Creating an AnnData object for single-cell or multi-sample analysis. Methods to analyze single-cell count matrix data, most relevant being software for single-cell RNA sequencing data, are readily applied to MuDPPI count matrices, whether at the protein or protein-pair level. Data exchange formats provide data sets in a format that can be imported by other software. The Python anndata package was used to export MuDPPI data for downstream analysis as an H5AD format file. The H5AD format is based on the hierarchical data format. Analyzing single-cell MuDPPI proteomics data. The Python scanpy library was used to explore MuDPPI data. Data sets saved as H5AD files are read by scanpy and used to analyze multiple samples that may be generated from bulk sample, from small numbers of cells, or from single cells. Analysis can be based on counts of individual proteins from individual A-side and B- side barcodes. Alternatively, analysis can be based on counts of interactions of protein pairs or on log-ratios of observed to expected interactions of protein pairs. It is understood that the foregoing detailed description and accompanying examples are merely illustrative and are not to be taken as limitations upon the scope of the disclosure, which is defined solely by the appended claims and their equivalents.
JHU Ref. No. C18037_P18037-02 Atty. Docket No. JHU-42990.601 Various changes and modifications to the disclosed embodiments will be apparent to those skilled in the art and may be made without departing from the spirit and scope thereof.